MAI Gateway - LLM Governance Deployment & Annual Managed Service
Sold by: 奇墨科技（广州）有限公司
It provides enterprises with deployment, implementation, configuration tuning, and one-year managed operation services for the MAI Gateway large model governance gateway, covering end-to-end delivery of multi-model access, cost control, intelligent routing and security audits.
Overview
This service provides deployment implementation and annual managed operation for the MAI Gateway enterprise-grade large model governance gateway deployed on Amazon EC2 (Amazon Web Services China).
MAI Gateway deployment implementation and annual hosting service
This service is aimed at enterprise customers who have purchased or are planning to use MAI Gateway, and provides complete delivery from environment construction and system integration to continuous operation for a period of one year. We transform the consumption of scattered, uncontrolled large model calls into enterprise-grade LLM infrastructure that is budgetable, attributable, auditable, and stable, making every token a manageable asset.
Service content
I. Deployment and implementation (one-time)
- Installation and initialization of the MAI Gateway gateway in the customer's Amazon Web Services environment (Amazon EC2)
- Multi-model aggregation access configuration: unified management of LLM APIs for multiple brands such as DeepSeek, Qwen, GLM, Kimi, MiniMax, Anthropic, OpenAI, and Google
- OpenAI compatible API interface, zero-code transformation of business systems
- Internal and external network isolation deployment architecture implemented, enterprise firewall unified exit, zero public network exposure
II. Governance system configuration (one-time)
- Building a 4-level hierarchical budget system: organization/department/project/key
- RBAC permission matrix configuration: four types of roles: employee, project leader, department administrator, and auditor
- Intelligent routing and failover policy configuration (default/ high availability/active/standby three modes)
- Token lifecycle key management, IP black-and-white list, access frequency limit, prompt injection protection, content filtering and sensitive data masking
- Annual escrow operation service (one year)
- MAI Gateway platform one-year license and operation guarantee
- 7x24 full-link monitoring: daily request volume, token consumption, interface failure rate (SLO ≤ 0.10%), abnormal events
- Second-level switching guarantee for upstream fluctuations, automatic downgrade and self-healing recovery
- Trace ID full link tracking and monthly cost attribution report (by department/project/user/token)
- 3-level consumption alarm (80%/95%/100%) and overrun processing
- Regular key rotation, visual cost dashboard operation support
- Release updates, security fixes, and technical support
Service object
Enterprise IT teams, finance and CFO teams, AI engineering teams, department managers, compliance and audit teams.
Delivery method
Services are delivered remotely + on-site (on demand) by the MAI Gateway team. The contract is for one year and can be renewed when it expires. The specific scope, cycle and price are subject to private quotation agreements.
Highlights
- Unified multi-model access, zero-code transformation — Integrate multi-brand models such as DeepSeek, Qwen, GLM, Anthropic, OpenAI, and Google through a single OpenAI compatible interface, decouple business logic from model vendors to eliminate vendor lock-up; allocate model visibility and calling rights by project or employee.
- Full-link cost governance — 4-level hierarchical budget (organization/department/project/key), millisecond real-time verification, 80%/95%/100% three-level alarms, month-end automatic attribution reports; visual dashboards present cost overviews, remaining budgets, and large consumers in real time, making each token budgetable, attributable, and optimizable.
- Production-level stability and security — Intelligent routing (default/high availability/active backup), automatic failover for upstream timeouts or interruptions, 7x24 monitoring and second-level switching; token lifecycle management includes regular rotation, IP black-and-white lists, prompt injection protection, content filtering, and sensitive data masking to meet security compliance requirements.
Details
Pricing
Custom pricing options
Pricing is based on your specific requirements and eligibility. To get a custom quote for your needs, request a private offer.
Legal
Content disclaimer
Vendors are responsible for their product descriptions and other product content. Amazon Web Services Marketplace China does not warrant that vendors' product descriptions or other product content are accurate, complete, reliable, current, or error-free.
Support
Vendor support
The MAI Gateway team provides enterprise customers with technical support during deployment implementation, configuration tuning, and annual operation. During the service period, 7x24 platform monitoring and alarm responses, as well as version updates and security fixes are provided. For assistance with product setup, troubleshooting, or general inquiries, please contact the MAI Gateway support team at aws@itqm.com ; support hours are 7x24, and we will respond to your support request within one business day.